BenchAtlas

Rankings / Unknown

Qwen3 Coder 30B A3b Instruct

released 2025-07-31

BenchAtlas Index

as of 2026-07-14
27.1
base configuration
rank #377
10 families · 5 categories · high

Benchmark evidence

13 results
agentic coding
τ²-Bench34.5 %independentArtificial Analysis
coding
LiveCodeBench v6
implementation=artificial-analysis
40.3 %independentArtificial Analysis
SciCode27.8 %independentArtificial Analysis
external indices
Intelligence Index v4.113.6 pointsindependentArtificial Analysis
AA Math Index29.0 pointsindependentArtificial Analysis
knowledge science
GPQA Diamond
implementation=artificial-analysis
51.6 %independentArtificial Analysis
Humanity's Last Exam
implementation=artificial-analysis
4.0 %independentArtificial Analysis
MMLU-Pro
implementation=artificial-analysis
70.6 %independentArtificial Analysis
long context instruction
AA-LCR29.0 %independentArtificial Analysis
IFBench32.6 %independentArtificial Analysis
reasoning math
AIME
implementation=artificial-analysis
29.7 %independentArtificial Analysis
AIME
year=2025 · implementation=artificial-analysis
29.0 %independentArtificial Analysis
MATH-500
implementation=artificial-analysis
89.3 %independentArtificial Analysis

Agent + model results

systems, not bare-model scores
agent + model EntroPO + Qwen3 Coder 30B A3b InstructSWE-bench Verified60.4 %unverifiedSWE-bench Leaderboard
agent + model EntroPO + Qwen3 Coder 30B A3b InstructSWE-bench Verified52.2 %unverifiedSWE-bench Leaderboard
agent + model OpenHands + Qwen3 Coder 30B A3b InstructSWE-bench Verified51.6 %unverifiedSWE-bench Leaderboard
agent + model EntroPO + Qwen3 Coder 30B A3b InstructSWE-bench Lite49.7 %unverifiedSWE-bench Leaderboard
agent + model EntroPO + Qwen3 Coder 30B A3b InstructSWE-bench Lite45.0 %unverifiedSWE-bench Leaderboard
agent + model Artificial Analysis harness + Qwen3 Coder 30B A3b InstructTerminal-Bench Hard15.2 %independentArtificial Analysis

These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.